关键词: 动态作业车间调度问题, 调度规则, 遗传规划, 仿真调度, 工艺路线相关准备时间

Abstract: The intelligent design of dispatching rules for dynamic job shop scheduling problem with separable sequence dependent setups was investigated by using the methods called genetic programming algorithm with the objective of minimizing the mean weighted tardiness.The procedure of genetic programming algorithm for designing dispatching rules was introduced in detail,the parameter setting of the algorithm for simulation experiments and the generation of the test case was presented as well.The scheduling results and the results of employed technique one way analysis of variances all demonstrated that the performance of rules generated by genetic programming was better than the selected benchmark rules.In addition,the homogeneity test was also adopted to test the sample means under different measures,and the results showed that the rules generated were robust under different scheduling environment.Overall,the analysis demonstrated that the approach based on genetic programming for intelligent design of dispatching rules aiming at specific scheduling environment was feasible and effective.In addition,the influences the model parameters on the performance of DRs were analyzed.The directions for future research were pointed out based on advantages and disadvantages of intelligence design of dispatching rules with genetic programming.

Key words: dynamic job shop scheduling problem, dispatching rules, genetic programming, simulation scheduling, sequence-dependent setup times